Applied Machine Learning for Early Detection of Retinal Toxicity

Hydroxychloroquine (HCQ) is an anti-inflammatory drug that is widely prescribed for a range of auto-immune disorders such as lupus and rheumatoid arthritis. An unwanted side effect of long-term use of HCQ is vision loss by retinal toxicity. If detected early, it could lead to early intervention to prevent vision loss and improve the quality of life for patients.
The project involves research on current machine learning approaches for the development of a system that would aid in the early detection of retinal toxicity. Current approaches involve qualitative interpretation of multifocal electroretinogram (mfERG) and optical coherence tomography (OCT) images by an expert. The project aims to develop a system that automates the interpretation of mfERG and OCT images to assist medical professionals in making an accurate diagnosis.

Faculty Supervisor:

Huaxiong Huang;Arvind Gupta

Student:

Faisal Habib

Partner:

Kensington Eye Institute

Discipline:

Computer science

Sector:

Health care and social assistance

University:

University of Toronto